Subway

​
Why we’re building the AI Gateway?

The AI development space is progressing at an exponential rate. We see it every day at Helicone—over 90 % of our users run 5 + LLMs in production, each with its own SDK, auth scheme, rate limits, and quirks. Keeping up today means rewriting integrations for every new model, managing a maze of API keys, engineering custom fallbacks for provider outages, and constantly tuning traffic for cost or compliance. Helicone AI Gateway is our answer. It is a lightweight Rust router inspired by NGINX that removes the integration tax so you can focus on shipping features.

​
What do you get with the AI Gateway?

One line. 100+ models

A unified interface for every LLM provider using familiar OpenAI syntax

Smart provider selection

Load balance to always hit the fastest, cheapest, or most reliable option

Spending controls

Rate limit to prevent runaway costs and usage abuse

Reduced latency

Cache responses to reduce costs and latency by up to 95%

Guaranteed uptime

Fallback to a different provider or model if the primary one is down

Centralized API keys

Store all API keys securely to end credential chaos and security risks

Simplified tracing

Monitor your AI workflows with built-in Helicone integration

One Click Deploy

Deploy to Fly.io, AWS, Porter, or any other platform in minutes

​
What sets our AI Gateway apart?

Get started instantly with our managed cloud offering:
  • Zero infrastructure - No servers to manage, no DevOps required
  • Instant setup - Start routing requests in 30 seconds
  • Auto-scaling - Handles any traffic volume automatically
  • Built-in monitoring - Integrated observability and analytics
  • 99.9% uptime SLA - Enterprise-grade reliability
  • Global edge network - Low latency worldwide

​
Self-Hosted

Built in Rust, the Gateway ships as one lightweight binary you can run anywhere:
  • Full control - Deploy on your infrastructure with complete customization
  • Sidecar-friendly - Drop into Docker, Kubernetes, bare-metal, or spawn as a subprocess
  • Built with Tower - Configurable middleware that leverages the Tower ecosystem
  • NGINX-style proxy - Local gateway to any provider, model, or region
  • Horizontally scalable - Run 1-N instances behind any load balancer
  • Open-source - Apache licensed, no vendor lock-in

​
Let’s get started!

Cloud Quickstart

Recommended - Get started in 30 seconds with zero setup

Self-Hosted Quickstart

Deploy on your own infrastructure